Precision M-S (Sum/Difference) Matrix Uses No Resistors
- Thread starter hejsan
- Start date
Help Support GroupDIY Audio Forum:
Handcrafted Tone
Well-known member
thats super cool :thumb:
I gotta try this when I get time (probably months from now)
I want to try a mastering technique I recently read about (I think it was in the last issue of tape op) - that is compressing the mid more than the sides. This should allow you to get louder mixes, and still have some dynamics and stereo image.
I gotta try this when I get time (probably months from now)
I want to try a mastering technique I recently read about (I think it was in the last issue of tape op) - that is compressing the mid more than the sides. This should allow you to get louder mixes, and still have some dynamics and stereo image.
enthalpystudios
Well-known member
those ll1517s!!!! i know, i'm going to build the same thing with mine ;]
been planning on it for a while now.
i'd be interested in trying out the THATencoder/decoder tho
been planning on it for a while now.
i'd be interested in trying out the THATencoder/decoder tho
Jim Zuehsow
Well-known member
Unfortunately the transformers Darius speaks of are rare as hen's teeth, especially here in the U.S., and are very expensive. Wayne's solution is simple and elegent at the same time. I remember when I first got an AKG C24, and the manual showed strapping a pair of A-67J's together to make a M/S matrix. It worked great, if you didn't mind a little HUM in your audio!I really do believe the circuit using the THAT components really would be every bit the equal of the finest German transformers from the past or present, and certainly better than the 685, or whatever the number is on the module I have been using.
Jim
Jim
Jim Zuehsow
Well-known member
I did have a really fine pair of transformers once and I think they were Malaki, or something like that. Supposedly manufactuered in the Black Forrest by elves. Anyway, they were super high quality and made specifically for M/S decoding. However, transformers, no matter how good, also bring their own set of "issues" to the table, and I could hear them. Not that they made things bad, but they added a certain character that wasn't there without them in the circuit. This may or may not be a good thing, depending what you're after. They certainly were noise free, large, and weighed a ton due to all the shielding. I finally sold them on ebay a few years ago to a mastering house that wanted them more than I did.
Jim
Jim
My WE 111 repeat coils work pretty well too.
I've also built the matrix out of the opamp I use in the EQ-2NV, and it sounds great, just a lot of parts.
I've built it out of OP-275's and matched R's too, not quite as sweet, but fine.
This approach looks real nice as well.
I've also built the matrix out of the opamp I use in the EQ-2NV, and it sounds great, just a lot of parts.
I've built it out of OP-275's and matched R's too, not quite as sweet, but fine.
This approach looks real nice as well.
111C's should be some of the best for M/S, given the tighter center tap design requirements. Make for a very large and manly M/S matrix. They do have a 'sound', but it's a sound I like in most cases. My Jensen iron is more transparent in comparison.
NewYorkDave
Well-known member
I moan under my breath whenever I see the 111C repeat coil mentioned. I could have gotten a few for free when I worked in radio... I really wish I'd grabbed 'em.
clintrubber
Well-known member
Nice all this, for just a few components good to have a 'fixed' version available as well.
Must have some TI-taste samples (INA134 ?), plan on using these for it.
Must have some TI-taste samples (INA134 ?), plan on using these for it.
Thnx MT.
I see the precision point.
BTW, one can precise-trim resistors (metal type) by hand. If you scratch a bit of color off resistor, or make a tiny scratch to surface with blade, resistance increases a bit. With this method I was able to finetune resistors to within 0.05% and better with some luck. It is easy to do it, and with some practice it is very fast to do.
Another thing regarding MS, for Cubase users, there are MS enc/dec plugins available (Voxengo even free), digital perfect precision, and you can insert that plug on your outputs, before going to console. Like this you can mix in MS, and then recording to another computer, you insert MS plugin again on Cubase input, monitoring with that plugin, and thus you can hear everything decoded in proper stereo, and master recording is also done in normal stereo. MS is thus used only through mixer and effects, for obvious mixing benefits like:
- easy level/width control for mid and side
- separate eq for mid and side
- separate compression
- modified console can give separate AUX sends for mid and side
- console eq irregularities do not influence stereo image
- many effects sound much better applied only to mid or only to side channel. (Applying fx only to side chan will put it out of phase in stereo though, which may or may not be desired)
On the negative side, MS mixing is a bit different, to say the least, and it takes some time to get used to it. Panning mono signals in stereo is best done within DAW, with stereo groups sent to MS mixer for levels and FX only. And instant stereo width control is just too addictive, going back to normal stereo mixing is like you'we lost whole lot of fun.
Just a few points from my experience with SM mixing (not mastering).
gnd
I see the precision point.
BTW, one can precise-trim resistors (metal type) by hand. If you scratch a bit of color off resistor, or make a tiny scratch to surface with blade, resistance increases a bit. With this method I was able to finetune resistors to within 0.05% and better with some luck. It is easy to do it, and with some practice it is very fast to do.
Another thing regarding MS, for Cubase users, there are MS enc/dec plugins available (Voxengo even free), digital perfect precision, and you can insert that plug on your outputs, before going to console. Like this you can mix in MS, and then recording to another computer, you insert MS plugin again on Cubase input, monitoring with that plugin, and thus you can hear everything decoded in proper stereo, and master recording is also done in normal stereo. MS is thus used only through mixer and effects, for obvious mixing benefits like:
- easy level/width control for mid and side
- separate eq for mid and side
- separate compression
- modified console can give separate AUX sends for mid and side
- console eq irregularities do not influence stereo image
- many effects sound much better applied only to mid or only to side channel. (Applying fx only to side chan will put it out of phase in stereo though, which may or may not be desired)
On the negative side, MS mixing is a bit different, to say the least, and it takes some time to get used to it. Panning mono signals in stereo is best done within DAW, with stereo groups sent to MS mixer for levels and FX only. And instant stereo width control is just too addictive, going back to normal stereo mixing is like you'we lost whole lot of fun.
Just a few points from my experience with SM mixing (not mastering).
gnd
I'd say MS is useful in more than just compression in mastering. EQ in MS is experience in itself. Say, boosting highs on Side chan, and taking them out on Mid chan a bit, widens perception of stereo field. Rolling off bass on Side chan focuses it on Mid chan, and makes it more defined for subwoofers.
Another problem with stereo mastering in subtle differences between signal chain of left and right channel. For example, EQ may sound a bit different on left than on right channel, and thus it smears stereo field (which may be good, or bad). In MS changes between both channels do not act on stereo field like that. I don't know how to put it best, but in MS stereo field stays more focused. But anyway, who am I to say, just my 2 cents...
I don't think that is the case. L+R linked sidechain is something else than MS compression.
...
I use it. But I'm like nobody. :grin: Maybe eq phase shifts can do something bad to signal, I cannot say. It sounds right to me, but you know, my ears, who knows... :roll: I'm no mastering wizard, I just do mastering (besides recording, mixing, producing, even soldering...), like many do, because we can, not because we really know how to do mastering. Would be nice to hear from some big mastering guys how they do things in MS.
Well.... It is easy to hear the mono mix with MS, you just pull down Side chan fader. All the way down is identical to mono button. But MS advantage is that you can 'drive' the Side fader to any position between full and zero. And of course, fun starts, when you actually drive side fader while recording final mix. That you cannot do in normal stereo mixing, but I find it essential to the magic of MS mixing. Automating (driving faders) in MS is whole another world than automating/driving in stereo.
But I'd say MS mixing is a privilege of small studios, where there is just one engineer/producer. If you hardwire your setup to MS, then anybody, who is not used to it, will be lost. But if you get used to it, it is great fun, full of advantages.
Sorry, cannot comment on GSSL, I'm not that fired up for it. I somehow prefer Waves SSL plugin, the digital heretic I am... :?
gnd
Andre
Well-known member
[quote author="mediatechnology"]
Would it be fair to say that M-S processing also has better mono fold-down/compatability? [/quote]
I think this would only be true if the source was recorded in M-S. As I understand it, a perfect encode and decode should produce the original signal with no change to Mono compatibility in the L/R domain. Of course, as gnd said, in the M/S domain you could just mute the side signal.
Would it be fair to say that M-S processing also has better mono fold-down/compatability? [/quote]
I think this would only be true if the source was recorded in M-S. As I understand it, a perfect encode and decode should produce the original signal with no change to Mono compatibility in the L/R domain. Of course, as gnd said, in the M/S domain you could just mute the side signal.
I’m really interested in this circuit for live monitoring purposes. Reading the posts and drawing from personal experience I would suggest the following feature set:
TRS Balanced I/O [designed so that it could be used balanced or unbalanced] and a four step width/mute switch with S channel mute, Narrow, Normal and Wide positions. [This would eliminate problems with a continuously variable pot and be satisfactory for live work.] For some purposes it would be implemented on the mixer insert points in my situation it would be inserted between the output of the multi-track and a stereo channel of the monitor mixer. I historically used M-S pairs in studio to capture back-up vocals, horn ensembles, Leslie cabinets, etc.
I hasten to point out that the circuit is useful as it sits and I do like the lower cost / smaller size / weight potential.
My application is for monitoring in a live setting. I want to record the mics in M-S mode for later adjustment of the stereo image in mixdown but be able to hear what is being captured. Ideally I’d like to have an M-S decoder matrix for each of the pairs that I might use. On a location capture I typically use at least 2 M-S pairs. Though I haven’t had occasion to, I can envision situations where I might use up to 4 M-S mics. Typically now I only monitor the M channel and leave the S channel for mixdown. Though it usually works, it does leave an uncomfortable margin for error.
I don’t have the stuff around for prototyping anymore so if anyone puts circuit boards together [or even better, a kit. Hint, hint.] I’d be very interested.
I’ve been searching and I haven’t been able to find anything like this out there.
TRS Balanced I/O [designed so that it could be used balanced or unbalanced] and a four step width/mute switch with S channel mute, Narrow, Normal and Wide positions. [This would eliminate problems with a continuously variable pot and be satisfactory for live work.] For some purposes it would be implemented on the mixer insert points in my situation it would be inserted between the output of the multi-track and a stereo channel of the monitor mixer. I historically used M-S pairs in studio to capture back-up vocals, horn ensembles, Leslie cabinets, etc.
I hasten to point out that the circuit is useful as it sits and I do like the lower cost / smaller size / weight potential.
My application is for monitoring in a live setting. I want to record the mics in M-S mode for later adjustment of the stereo image in mixdown but be able to hear what is being captured. Ideally I’d like to have an M-S decoder matrix for each of the pairs that I might use. On a location capture I typically use at least 2 M-S pairs. Though I haven’t had occasion to, I can envision situations where I might use up to 4 M-S mics. Typically now I only monitor the M channel and leave the S channel for mixdown. Though it usually works, it does leave an uncomfortable margin for error.
I don’t have the stuff around for prototyping anymore so if anyone puts circuit boards together [or even better, a kit. Hint, hint.] I’d be very interested.
I’ve been searching and I haven’t been able to find anything like this out there.
bobschwenkler
Well-known member
Hey All,
The first post of this thread seems to be missing. Does anyone have a copy of the schematic in question?
The first post of this thread seems to be missing. Does anyone have a copy of the schematic in question?
